StanCon 2018

StanCon 2018

Three days of talks, open discussions, and statistical modeling.

Jan 10-12, 2018   •   Asilomar Conference Center, Pacific Grove, California, USA


Slides, code, and other materials from presentations at StanCon 2018 are now available at our stancon_talks repository:

Links are also provided from the conference schedule below.



Sponsors

Thanks to these fine folks for sponsoring StanCon2018


Scholarships

Scholarshipes are closed.


Wearable Poster Session

We will accept poster submissions on a rolling basis until January 9th. One page exclusive of references is the desired format but anything that gives us enough information to make a decision is fine. We will accept/reject within min(48 hours,time before reception - 1 hr). Send to stancon2018@mc-stan.org or just come talk to Breck at the conference. The only somewhat odd requirement is that your poster must be “wearable” to the 5pm reception where you will be a walking presentation. Great way to network, signboard supplies will be available so you need only have sheets of paper which can be attached to signboard material.

Location

Asilomar Conference Grounds, Pacific Grove, California

StanCon 2018 will take place at the Asilomar Conference Grounds, a National Historic Landmark on the Monterey Peninsula right on the beach.

Asilomar is about 120 miles south of San Francisco and 75 miles south of San Jose. Shuttle service is $60 each way from San Francisco airport (3 hour trip) and $50 each way from San Jose airport (2 hour trip) (https://www.montereyairbus.com). Shuttles run approximately every 1.5 hours. Car rental is also an option and showing rates less than $30/day as of Oct 29th.

Invitation Letters for US visa

If you require an invitation letter for visa or other purposes please fill out this form. We will email you a letter within 48 hours. You must be registered or a scholarship awardee to get the letter.

Registration Information

IMPORTANT: Tell us your T-shirt size and tutorial/activity preferences on our SWAG form.

If you are staying at the conference facility, Asilomar, then registration for the conference is handled as a part of the room booking process. Register/book room at Asilomar. More accomidation information below.

Offsite registration is done via EventBrite. Please do not use this to register and then stay at Asilomar with some cheap deal. Our room rate covers a bunch of conference expenses. See below for meal costs.

Early Registration (ends Dec 15)

Student Academic Industry
$150 $250 $350


Regular Registration

Student Academic Industry
$225 $375 $525


We are offering $50 discounted registration to anyone identifying with as a member of an underrepresented community, including but not limited to gender, ethnicity, and orientation.

Accommodation

Onsite Accommodations (will also collect payment for StanCon2018 registration)

Book accommodations!

Accommodations at Asilomar include meals. Cost for 4 days, 3 nights (Jan 9-12):

  • Single accommodations: $899.00
  • Double accommodations: $605.93

These costs do not include an additional $20 processing fee that will be charged when you check into your accommodation at Asilomar.

Offsite Accommodations

For those attending the conference, but not staying at Asilomar, meals can be purchased a la carte:

  • Breakfast: $19.81
  • Lunch: $25.64
  • Dinner: $39.87

Speakers and Schedule

Going to be a pressure cooker folks. Subject to change without notice.

Merrill Hall is where StanCon2018 happens. Exceptions are meals, after dinner activities and some tutorials (Scripps Hall).

Wednesday, Jan 10

Breakfast 7:30-9:00am

Tutorials/Classes 8:00-10:00am

Break 10:00-10:40am

Invited Talk: Predictive Information Criteria in Hierarchical Bayesian Models for Clustered Data. Sophia Rabe-Hesketh and Daniel Furr (U California, Berkely) 10:40-11:30am (Video, Slides)

  • Opening Comments and Information you Need, Breck Baldwin 11:30-11:40am

  • Does the New York City Police Department rely on quotas? Jonathan Auerbach (Columbia U) 11:40-12:00am Video (Notebook and slides)
  • Diagnosing Alzheimer’s the Bayesian Way. Arya A. Pourzanjani, Benjamin B. Bales, Linda R. Petzold, Michael Harrington (UC Santa Barbara) 12:00am-12:20pm Video (Notebook and slides)
  • Joint longitudinal and time-to-event models via Stan. Sam Brilleman, Michael Crowther, Margarita Moreno-Betancur, Jacqueline Buros Novik, Rory Wolfe (Monash U, Columbia U) 12:20-12:40pm Video(Notebook and slides)

Lunch 12:40-2:00pm

  • ScalaStan. Joe Wingbermuehle (Cibo Technologies) 2:00-2:20pm Video (Slides)
  • A tutorial on Hidden Markov Models using Stan. Luis Damiano, Brian Peterson, Michael Weylandt 2:20-2:40pm Video (Notebook and slides)
  • Student Ornstein-Uhlenbeck models served three ways (with applications for population dynamics data). Aaron Goodman (Stanford U) 2:40-3:00pm (Notebook and slides)
  • Spatial models in Stan: intrinsic auto-regressive models for areal data. Mitzi Morris (Columbia U) 3:00-3:20pm (Slides, Video Stan case study)

Break 3:20-4:00pm

Invited Talk: Stan Applications in Physics: Testing Quantum Mechanics and Modeling Neutrino Masses, Talia Weiss (MIT) 4:00-4:50pm (Slides)

Reception 5:00-6:00pm

Dinner 6:00-7pm

After Dinner 7-9pm

  • Build and fly your own R/C airplane Not any more folks, FAA has reinstated R/C aircraft restrictions. We will be doing air powered rockets with illumination. Don’t worry, they are super fun and go really high.
  • S’mores by the outdoor fire place

Thursday, Jan 11

Breakfast 7:30-9:00am

Tutorials/Classes

Break 10:00-10:40am

Invited Talk: Forecasting at Scale: How and why we developed Prophet for forecasting at Facebook, Sean Taylor and Ben Letham (Facebook) 10:40-11:30am Video (Details)

  • SlicStan: a blockless Stan-like language. Maria I. Gorinova, Andrew D. Gordon, Charles Sutton (U of Edinburgh) 11:30-11:50am Video (Notebook and slides)
  • Introducing idealstan, an R package for ideal point modeling with Stan. Robert Kubinec (U of Virginia) 11:50am-12:10pm Video (Notebook and slides)
  • A brief history of Stan. Daniel Lee (Generable) 12:10-12:30pm Video (Slides)

Lunch 12:30-1:30pm

  • Computing steady states with Stan’s nonlinear algebraic solver. Charles C. Margossian (Metrum, Columbia U) 1:30-1:50pm Video (Notebook and slides)
  • Bayesian estimation of mechanical elastic constants. Ben Bales, Brent Goodlet, Tresa Pollock, Linda Petzold (UC Santa Barbara) 1:50-2:10pm Video (Notebook and slides)
  • Mystery Talk, Mystery Person, Regularized Horseshoe, Aki Vehtari Video (Aalto U) 2:10-2:30

Modeling/Data Session + Classes 2:30-4:10pm

Invited Talk: Stan applications in Human Genetics: Prioritizing genetic mutations that protect individuals from human disease, Manuel Rivas (Stanford U) 4:00-4:50pm Videos (Slides)

Reception 5:00-6:00pm

Dinner 6:00-7:00pm

After Dinner 7:00-9:00pm

  • Non-snobby Blind Wine Tasting

Friday, Jan 12

Breakfast 7:30-9:00am

Tutorials/Classes

Workshop on Model Selection: Aki Vehtari (Aalto U) 9:00-10:00am Video (Materials)

Break 10:00-10:40am

Invited Talk: Susan Holmes (Stanford U) 10:40-11:30am Video (Slides)

  • Aggregate random coefficients logit — a generative approach. Jim Savage, Shoshana Vasserman 11:30-11:50am Video (Notebook and slides)
  • The threshold test: Testing for racial bias in vehicle searches by police. Camelia Simoiu, Sam Corbett-Davies, Sharad Goel, Emma Pierson (Stanford U) 11:50am-12:10pm Video (Notebook and slides)
  • Assessing the safety of Rosiglitazone for the treatment of type II diabetes. Konstantinos Vamvourellis, K. Kalogeropoulos, L. Phillips (London School of Economics and Political Science) 12:10-12:30pm (Video, Notebook and slides)

Lunch 12:30-1:30pm

  • Causal inference with the g-formula in Stan. Leah Comment (Harvard U) 1:30-1:50pm Video (Notebook and slides)
  • Bayesian estimation of ETAS models with Rstan. Fausto Fabian Crespo Fernandez (Universidad San Francisco de Quito) 1:50-2:10pm Video (Notebook and slides)

Invited Talk: Some problems I’d like to solve in Stan, and what we’ll need to do to get there, Andrew Gelman 2:10-3:00 (Columbia U) (Video)

Classes/Tutorials

We have tutorials that start at the crack of 7am for those desiring further edification beyond the program–these do not run in parallel to the main session but do run parallel to each other–rooms to be determined (Merrill or Scripps):

  • Introduction to Stan: Know how to program? Know basic statistics? Curious about Bayesian analysis and Stan? This is the course for you. Hands on, focused and an excellent way to get started working in Stan. Wed-Fri 8-10am, Thurs 2:30-4:10pm. Jonah Sol Gabry, Mitzi Morris, Sean Talts.

  • Executive decision making the Bayesian way: This is for non-technical managers and technical folks who need to communicate with managers to learn the core of decision making under uncertainty. Wed 8-9am, Thurs 3:30-4:10, Friday 8-9am. Jonathan Auerback, Eric Novik.

  • Advanced Hierarchical Models in Stan: The hard stuff. Ben Goodrich. 9-10am Wed, 7-8am Thurs, 7-8am Friday.

  • Model assessment, selection and inference after model selection The tutorial covers cross-validation, reference predictive and projection predictive approaches for model assessment, selection and inference after model selection. I discuss when cross-validation is useful and when we can do better than cross-validation (and why not to use WAIC). The tutorial is accommopanied with R notebooks using rstanarm, bayesplot, loo, and projpred packages. You can bring also your own data and model to get recommendations for model assessment and selection. Aki Vehtari, Thurs 8-9am

  • How to develop for Stan at the C++ level: Overview of Stan C++ architecture and build/development process for contributors. Charles Margossian, 9-10am Thurs.

  • A Dive into Stan’s C++ Model Concept This will be a discussion partially driven by the participants. We will cover things like: - how the Stan language is translated to C++ - overview of the C++ model concept that’s generated - how to use the math library for autodiff - how to use the generated C++ model concept to write an inference algorithm in C++ - opportunities in efficient computing once the C++ is generated

      Please be comfortable with C++ or at least be willing to stare at blocks of ugly code. Please have CmdStan v2.17.1 installed. Daniel Lee, 9-10am Friday.
    

Activities

Learning is fun but we anticipate that blowing off a little steam will be called for.

R/C Airplanes Now Air Powered Rockets

R/C Airplanes are cancled due to FAA regulation changes in December. In place we will have build your own air powered rockets and they will be carrying night lights. After dinner on Wed eve we will provide designs and building materials to create your own rocket. It takes about 15 minutes if you are quick and we will have glow sticks to put on them.

S’mores By the Bonfire

Pretty much self explanatory. Wed eve.

Snob-free Blind Wine Tasting

By day 2 you will have gotten to know your fellow attendees so some social adventure is called for. This activity has proved wildly successful at DARPA conferences and they invented the internet so it can’t be all bad. Participants taste wines without knowing what they are.

Code of Conduct

In order to facilitate a welcoming environment for all attendees, StanCon 2018 will enforce a code of conduct.

StanCon 2018 Organizing Committee

If you have any questions regarding StanCon, please email us at stancon@mc-stan.org.

StanCon is organized by a volunteer community

  • Breck Baldwin (Columbia University)
  • Jonah Gabry (Columbia University)
  • Daniel Lee (Generable, Inc)
  • Eric Novik (Generable, Inc)
  • Megan Shabram (NASA Ames)
  • Rob Trangucci (University of Michigan)
  • Lizzie Wolkovich (Harvard University)